Professional Documents
Culture Documents
org
doi:10.14355/jitae.2013.0204.10
ITMediatedDevelopmentofSustainable
AgricultureSystems
TowardaDataDrivenCitizenScience
MasatoshiFunabashi*1
SonyComputerScienceLaboratories,inc.
Takanawamusebldg.3F,31413,HigashiGotanda,Shinagawaku,Tokyo,1410022,Japan
*1
masa_funabashi@csl.sony.co.jp
Abstract
Sustainable farming is facing an informatization challenge
on the complexity of ecosystem dynamics. We introduced
collective pervasive learning using webbased database on
biodiversity and farming knowledge in polyculture system.
Thesharingofknowledgeanddataisshowntobeeffective
bothinunderstandingthebehaviourofecosystemsandthe
management of vegetation portfolio in relation to local
environment.
Keywords
Sustainable Agriculture; Pervasive Learning; ComputerAided
Assessment;EnvironmentalManagementSystems
Introduction
Agricultureisoneofthecentralissuesinsustainability
of both human society and planetary environment
[Barnosky 2012]. Several international organizations
areraisingalertinscientificreportstopreventfurther
biodiversity loss and convert our socioecological
systems to a sustainable design (i.e., [IAASTD 2009]
[MEAB2005]).Theargumentsmutualyemphasizethe
importance of small holder farmers whose
contributiontolocaleconomy,nutritiondiversity,and
environmental protection is underexploited (i.e.,
[HLPE2013]).
IT based enhancement of small scale agriculture is
recently emerging, and further develppoment is
needed especially for the food security in developing
countries such as India and China, with a large
population of smallfarm holders as substantial
suppliers[Balga2013][Zhang2013].
Small farms usually practice with locally diversified
methods, depending on the ecological, geographical
and social contexts.The optimization of conventional
largescalefarmingsuchasprecisionagricultureisnot
necessary enough to treat such diversity of both
environment and method (i.e. Traditional small
practices do not usually involve machinery). Rather,
179
www.jitae.orgJournalofInformationTechnologyandApplicationinEducationVol.2Iss.4,Dec.2013
farmingsite,andsoon.
biodiversityandsharingofknowledge,listsofspecies
taxonomy classification (32,567 plant species and
30,277 insect species from opensource database),
associationstrategyofplants,productivityresults,etc.
Synecoculture Experiment
SynecocultureFarmingMethod
Weappliedthemethodologyofpervasivelearningfor
the management of Synecoculture method in Japan.
Synecoculture is a smallscale polyculture system,
producing a wide variety of vegetables, herbs, fruits,
etc [Funabashi 2011]. Its challenge is tointelligently
manage biodiversity of the field to profit from a
variety of ecosystems function such as productivity,
resilience, product quality, etc. Conventional
methodology that produces environmental load is
avoided, such as tillage, fertilizer, chemicals and
machinery. Fig.1 and 2 show the experiment site of
Synecoculture. It is based on a dense polyculture of
edible plants, each species competing with others to
form their niche in a state of ecological optimum. In
such practice, the sharing of knowledge on
biodiversity and farming strategy such as vegetation
portfoliobecomeessentialforthemanagement.
Left:Photoofthefield.Right:Schematicsegmentationofproductive
surface(greens)andfruittrees(oranges)
FIG.1OVERVIEWOFSYNECOCULTUREFIELD
Left:Photoof4 productivesurface.Right:Schematic
segmentationofdifferentkindsofvegetables(13species).
Top:Searchsequenceofspeciesnamefromrelatedtaxonomyand
metadata.Usercanfindspeciesnamefromotherinformation.
Bottom:Autofillingofrelatedtaxonomicalinformationandmeta
data.Registeredspeciesisautomaticallyrelatedwithtaxonomyand
metadataspecifictospecies.
FIG.2CLOSEVIEWOFSYNECOCULTUREFIELD
FIG.3DATABASESTRUCTUREFORTHESUPPORTOF
PERVASIVELEARNINGINSYNECOCULTUE
ComputerAidedAssessmentandManagement
DatabaseStructure
180
JournalofInformationTechnologyandApplicationinEducationVol.2Iss.4,Dec.2013www.jitae.org
Greennodes:Plants,Rednodes:Insectes,Arrows:Interactionsbased
onliteratures
FIG.5EXAMPLEOFRECONSTRUCTEDECOLOGICAL
NETWORKBETWEENPLANTSANDINSECTSIN
SYNECOCULTUREFIELD
DatabasedrivenEnhancementofBiodiversity
Candidatesofmanagementvariables(nodes)andtheir
dependenciesindecisionmaking(links)aredepictedasanetwork.
FIG.EXAMPLEOFMANAGEMENTMODELKNOWLEDGE
STRUCTURE
FIG.6TOP:EXAMPLESOF27RHOPALOCERASPECIES
RECORDEDINSYNECOCULTUREFIELDS.BOTTOM:
PREDICTIONANDRESULTOFRHOPALOCERAVISITIN
SYNECOCULTUREFIELDS.POPULARITYRANKS(04)AND
PREDICTIONRATE(REDBAR)WEREDEFINEDUSINGLOCAL
ECOLOGICALRECORDSANDACTUALVEGETATIONDATA.
THERESULTS(BLUEANDGREENBARS)ARETHE
PERCENTAGEOFTHEACCURACYOFPREDICTIONINEACH
SUBSETOFFIELDS.
181
www.jitae.orgJournalofInformationTechnologyandApplicationinEducationVol.2Iss.4,Dec.2013
attainsustainabilitygoals[Ehrlich2012].
ACKNOWLEDGMENT
Conclusions
(2011).
HLPE.Investinginsmallholderagricultureforfoodsecurity.
MasatoshiFunabashi
182